论文部分内容阅读
对产品族配置设计案例推理方法中产品属性不确定引起的问题进行了研究。为同时考虑产品属性的重要性及顾客偏好,定义加权Euclid距离为目标案例与源案例的相似度。用区间数表征产品属性的不确定性,利用蒙特卡洛模拟(Monte Carle Simulation,MC)方法计算点与区间、区间与区间的Euclid距离,得到相似度概率分布的特征值(均值和方差)。再利用集对分析(Set Pair Analysis,SPA)判定产品属性不确定性对案例检索结果的影响,并最终确定满足客户要求的产品族。油锯配置设计实例应用表明,MC-SPA方法组合了MC方法的直观性、计算的精确性及SPA方法处理不确定性问题的优势,案例检索灵敏度高,检索结果稳定可靠。
The problems caused by the uncertainty of product attributes in product case configuration design case inference method were studied. To consider both the importance of product attributes and customer preference, the weighted Euclid distance is defined as the similarity between the target case and the source case. The interval number is used to characterize the uncertainty of product attributes. Monte Carlo simulation (MC) is used to calculate the Euclid distance between points and intervals, intervals and intervals, and the eigenvalues (mean and variance) of the probability distributions are obtained. Set Pair Analysis (SPA) is then used to determine the impact of product attribute uncertainty on the case retrieval results, and ultimately determine the product family that meets customer requirements. The application of the design example of the saw shows that the MC-SPA method combines the intuition of the MC method, the accuracy of the calculation and the advantages of the SPA method in handling the uncertainties. The sensitivity of the case retrieval is high and the retrieval result is stable and reliable.